6 ms·
That paper is kind of weird -- linear PDEs are solved by convolutions (yay for Green's functions), so using a CNN seems a bit, well, overkill and obvious?
by santaclaus 10y ago
That paper is kind of weird -- linear PDEs are solved by convolutions (yay for Green's functions), so using a CNN seems a bit, well, overkill and obvious?